Studies suggesting that increased trauma center (TC) patient volume leads to improved patient outcomes have produced inconsistent results. It is possible that the combination of TC patient volume and TC resources has a greater influence over outcomes than TC volume alone. Patients admitted to 131 TCs from 1994 to 2002 were included. TC density was calculated as the ratio of TC adult beds to patient volume and categorized into quartiles. Adjusted risk ratios (RRs) and 95 per cent confidence intervals (CIs) compared mortality of groups (Q1 [low density]-Q4 [high density]). Relative to Q4 TCs, for penetrating injuries with ISS < 15, elevated mortality was seen for Q2 (low-moderate density) TCs (RR 1.47, CI 1.09-1.99) and Q3 (high-moderate density) TCs (RR 1.81, CI 1.36-2.41); but for ISS > or = 15, mortality was lowest for Q1 (RR 0.78, CI 0.68-0.89). For blunt injury, mortality was similar for low- and high-density TCs but elevated for low-moderate- and high-moderate-density TCs. Mortality at low-density TCs was decreased overall and for penetrating injuries specifically but similar for blunt injuries when compared with high-density TCs. Enhanced resources measurements will clarify the relationship between patient volume, resources, and mortality.